I've a work project that requires a stat test between Product A and Produce B performance ratings (On a scale of 1 to 5 how well did Product A / B perform?)
We are treating the data as interval scaled.
The samples partially overlap. e.g. Product A has n=500; Product B has n=500...